Authors: Mark Kevin Salloway, Zheng Jye Ling, Qian Yang, Hwee Yeong Lew, Xiaodong Deng, Kee Seng Chia, Yik Ying Teo, E-Shyong Tai, Chuen Seng Tan
The advent of the electronic medical records (EMRs) data for large scale application of data science and analytics in healthcare has opened up the possibilities to improve human lives through empirical evidence and knowledge discovery. In particular, the potential of using operational data as a data source allows academic medical centers with EMRs to make profound impact through better use of resources, drive re-engineering of processes and workflows, and identify areas of unmet need. Although retrospective EMR data have limitations which impact analytics, a step forward can begin with a plan to develop an infrastructure utilising such data. This facilitates the building up of capabilities and resources for handling large and complex data analytics in the future. This paper describes some of the processes to consider when preparing to house such datasets for research purposes, especially the securing of personal identifiers. Taking small steps with retrospective EMR data can initiate the start of information flow for healthcare data analytics, while concurrently developing the expertise useful for future EMR systems that cater to the needs of public health.
Keywords: electronic medical records; healthcare; public health; data analytics; data protection